Pass job time range / duration to calls to /controller/rest/applications/application_name/nodes

Currently Dexter assumes you want all nodes returned in node reports, even inactive ones that will eventually be purged. The request is to enhance Dexter to pass through the job timerange to the timeRange parameter of the calls made to /controller/rest/applications/application_name/nodes API via controllerApi.GetAPMNodes(), ensuring that only 'live' nodes during that time range are returned, not stale nodes or nodes on which the clock has started for purging.

Start by locating controllerApi.GetAPMNodes() and the callers that know the job time range. Trace how requests are made to the applications/{application_name}/nodes API, then verify that the job range is passed as timeRange and that returned nodes are limited to those live during that range.

Written by the indexing model from the issue text.